rocksdb/tools/ldb_cmd_test.cc

1265 lines
40 KiB
C++
Raw Normal View History

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
#include "rocksdb/utilities/ldb_cmd.h"
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
#include <cinttypes>
#include "db/db_test_util.h"
#include "db/version_edit.h"
#include "db/version_set.h"
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
2019-12-13 22:47:08 +00:00
#include "env/composite_env_wrapper.h"
#include "file/filename.h"
#include "port/stack_trace.h"
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
#include "rocksdb/advanced_options.h"
#include "rocksdb/comparator.h"
#include "rocksdb/convenience.h"
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
#include "rocksdb/db.h"
#include "rocksdb/file_checksum.h"
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
#include "rocksdb/file_system.h"
#include "rocksdb/utilities/options_util.h"
#include "test_util/sync_point.h"
#include "test_util/testharness.h"
#include "test_util/testutil.h"
#include "util/file_checksum_helper.h"
#include "util/random.h"
using std::map;
using std::string;
using std::vector;
namespace ROCKSDB_NAMESPACE {
class LdbCmdTest : public testing::Test {
public:
LdbCmdTest() : testing::Test() {}
Env* TryLoadCustomOrDefaultEnv() {
Env* env = Env::Default();
EXPECT_OK(test::CreateEnvFromSystem(ConfigOptions(), &env, &env_guard_));
return env;
}
private:
std::shared_ptr<Env> env_guard_;
};
TEST_F(LdbCmdTest, HelpAndVersion) {
Options o;
o.env = TryLoadCustomOrDefaultEnv();
LDBOptions lo;
static const char* help[] = {"./ldb", "--help"};
ASSERT_EQ(0, LDBCommandRunner::RunCommand(2, help, o, lo, nullptr));
static const char* version[] = {"./ldb", "--version"};
ASSERT_EQ(0, LDBCommandRunner::RunCommand(2, version, o, lo, nullptr));
static const char* bad[] = {"./ldb", "--not_an_option"};
ASSERT_NE(0, LDBCommandRunner::RunCommand(2, bad, o, lo, nullptr));
}
TEST_F(LdbCmdTest, HexToString) {
// map input to expected outputs.
// odd number of "hex" half bytes doesn't make sense
map<string, vector<int>> inputMap = {
{"0x07", {7}}, {"0x5050", {80, 80}}, {"0xFF", {-1}},
{"0x1234", {18, 52}}, {"0xaaAbAC", {-86, -85, -84}}, {"0x1203", {18, 3}},
};
for (const auto& inPair : inputMap) {
auto actual = ROCKSDB_NAMESPACE::LDBCommand::HexToString(inPair.first);
auto expected = inPair.second;
for (unsigned int i = 0; i < actual.length(); i++) {
EXPECT_EQ(expected[i], static_cast<int>((signed char)actual[i]));
}
auto reverse = ROCKSDB_NAMESPACE::LDBCommand::StringToHex(actual);
EXPECT_STRCASEEQ(inPair.first.c_str(), reverse.c_str());
}
}
TEST_F(LdbCmdTest, HexToStringBadInputs) {
const vector<string> badInputs = {
"0xZZ", "123", "0xx5", "0x111G", "0x123", "Ox12", "0xT", "0x1Q1",
};
for (const auto& badInput : badInputs) {
try {
ROCKSDB_NAMESPACE::LDBCommand::HexToString(badInput);
std::cerr << "Should fail on bad hex value: " << badInput << "\n";
FAIL();
} catch (...) {
}
}
}
TEST_F(LdbCmdTest, MemEnv) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
for (int i = 0; i < 100; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
ASSERT_OK(db->Put(wopts, buf, buf));
}
FlushOptions fopts;
fopts.wait = true;
ASSERT_OK(db->Flush(fopts));
delete db;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "dump_live_files";
char* argv[] = {arg1, arg2, arg3};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(3, argv, opts, LDBOptions(), nullptr));
}
class FileChecksumTestHelper {
private:
Options options_;
DB* db_;
std::string dbname_;
Status VerifyChecksum(LiveFileMetaData& file_meta) {
std::string cur_checksum;
std::string checksum_func_name;
Status s;
EnvOptions soptions;
std::unique_ptr<SequentialFile> file_reader;
std::string file_path = dbname_ + "/" + file_meta.name;
s = options_.env->NewSequentialFile(file_path, &file_reader, soptions);
if (!s.ok()) {
return s;
}
std::unique_ptr<char[]> scratch(new char[2048]);
Slice result;
FileChecksumGenFactory* file_checksum_gen_factory =
options_.file_checksum_gen_factory.get();
if (file_checksum_gen_factory == nullptr) {
cur_checksum = kUnknownFileChecksum;
checksum_func_name = kUnknownFileChecksumFuncName;
} else {
FileChecksumGenContext gen_context;
gen_context.file_name = file_meta.name;
std::unique_ptr<FileChecksumGenerator> file_checksum_gen =
file_checksum_gen_factory->CreateFileChecksumGenerator(gen_context);
checksum_func_name = file_checksum_gen->Name();
s = file_reader->Read(2048, &result, scratch.get());
if (!s.ok()) {
return s;
}
while (result.size() != 0) {
file_checksum_gen->Update(scratch.get(), result.size());
s = file_reader->Read(2048, &result, scratch.get());
if (!s.ok()) {
return s;
}
}
file_checksum_gen->Finalize();
cur_checksum = file_checksum_gen->GetChecksum();
}
std::string stored_checksum = file_meta.file_checksum;
std::string stored_checksum_func_name = file_meta.file_checksum_func_name;
if ((cur_checksum != stored_checksum) ||
(checksum_func_name != stored_checksum_func_name)) {
return Status::Corruption(
"Checksum does not match! The file: " + file_meta.name +
", checksum name: " + stored_checksum_func_name + " and checksum " +
stored_checksum + ". However, expected checksum name: " +
checksum_func_name + " and checksum " + cur_checksum);
}
return Status::OK();
}
public:
FileChecksumTestHelper(Options& options, DB* db, std::string db_name)
: options_(options), db_(db), dbname_(db_name) {}
~FileChecksumTestHelper() = default;
// Verify the checksum information in Manifest.
Status VerifyChecksumInManifest(
const std::vector<LiveFileMetaData>& live_files) {
// Step 1: verify if the dbname_ is correct
if (dbname_.back() != '/') {
dbname_.append("/");
}
// Step 2, get the the checksum information by recovering the VersionSet
// from Manifest.
std::unique_ptr<FileChecksumList> checksum_list(NewFileChecksumList());
EnvOptions sopt;
std::shared_ptr<Cache> tc(NewLRUCache(options_.max_open_files - 10,
options_.table_cache_numshardbits));
options_.db_paths.emplace_back(dbname_, 0);
options_.num_levels = 64;
WriteController wc(options_.delayed_write_rate);
WriteBufferManager wb(options_.db_write_buffer_size);
ImmutableDBOptions immutable_db_options(options_);
VersionSet versions(dbname_, &immutable_db_options, sopt, tc.get(), &wb,
Make OffpeakTimeInfo available in VersionSet (#12018) Summary: As mentioned in https://github.com/facebook/rocksdb/issues/11893, we are going to use the offpeak time information to pre-process TTL-based compactions. To do so, we need to access `daily_offpeak_time_utc` in `VersionStorageInfo::ComputeCompactionScore()` where we pick the files to compact. This PR is to make the offpeak time information available at the time of compaction-scoring. We are not changing any compaction scoring logic just yet. Will follow up in a separate PR. There were two ways to achieve what we want. 1. Make `MutableDBOptions` available in `ColumnFamilyData` and `ComputeCompactionScore()` take `MutableDBOptions` along with `ImmutableOptions` and `MutableCFOptions`. 2. Make `daily_offpeak_time_utc` and `IsNowOffpeak()` available in `VersionStorageInfo`. We chose the latter as it involves smaller changes. This change includes the following - Introduction of `OffpeakTimeInfo` and `IsNowOffpeak()` has been moved from `MutableDBOptions` - `OffpeakTimeInfo` added to `VersionSet` and it can be set during construction and by `ChangeOffpeakTimeInfo()` - During `SetDBOptions()`, if offpeak time info needs to change, it calls `MaybeScheduleFlushOrCompaction()` to re-compute compaction scores and process compactions as needed Pull Request resolved: https://github.com/facebook/rocksdb/pull/12018 Test Plan: - `DBOptionsTest::OffpeakTimes` changed to include checks for `MaybeScheduleFlushOrCompaction()` calls and `VersionSet`'s OffpeakTimeInfo value change during `SetDBOptions()`. - `VersionSetTest::OffpeakTimeInfoTest` added to test `ChangeOffpeakTimeInfo()`. `IsNowOffpeak()` tests moved from `DBOptionsTest::OffpeakTimes` Reviewed By: pdillinger Differential Revision: D50723881 Pulled By: jaykorean fbshipit-source-id: 3cff0291936f3729c0e9c7750834b9378fb435f6
2023-10-27 22:56:48 +00:00
&wc, nullptr, nullptr, "", "",
options_.daily_offpeak_time_utc, nullptr,
/*read_only=*/false);
std::vector<std::string> cf_name_list;
Status s;
s = versions.ListColumnFamilies(&cf_name_list, dbname_,
Simplify migration to FileSystem API (#6552) Summary: The current Env/FileSystem API separation has a couple of issues - 1. It requires the user to specify 2 options - ```Options::env``` and ```Options::file_system``` - which means they have to make code changes to benefit from the new APIs. Furthermore, there is a risk of accessing the same APIs in two different ways, through Env in the old way and through FileSystem in the new way. The two may not always match, for example, if env is ```PosixEnv``` and FileSystem is a custom implementation. Any stray RocksDB calls to env will use the ```PosixEnv``` implementation rather than the file_system implementation. 2. There needs to be a simple way for the FileSystem developer to instantiate an Env for backward compatibility purposes. This PR solves the above issues and simplifies the migration in the following ways - 1. Embed a shared_ptr to the ```FileSystem``` in the ```Env```, and remove ```Options::file_system``` as a configurable option. This way, no code changes will be required in application code to benefit from the new API. The default Env constructor uses a ```LegacyFileSystemWrapper``` as the embedded ```FileSystem```. 1a. - This also makes it more robust by ensuring that even if RocksDB has some stray calls to Env APIs rather than FileSystem, they will go through the same object and thus there is no risk of getting out of sync. 2. Provide a ```NewCompositeEnv()``` API that can be used to construct a PosixEnv with a custom FileSystem implementation. This eliminates an indirection to call Env APIs, and relieves the FileSystem developer of the burden of having to implement wrappers for the Env APIs. 3. Add a couple of missing FileSystem APIs - ```SanitizeEnvOptions()``` and ```NewLogger()``` Tests: 1. New unit tests 2. make check and make asan_check Pull Request resolved: https://github.com/facebook/rocksdb/pull/6552 Reviewed By: riversand963 Differential Revision: D20592038 Pulled By: anand1976 fbshipit-source-id: c3801ad4153f96d21d5a3ae26c92ba454d1bf1f7
2020-03-24 04:50:42 +00:00
immutable_db_options.fs.get());
if (s.ok()) {
std::vector<ColumnFamilyDescriptor> cf_list;
for (const auto& name : cf_name_list) {
fprintf(stdout, "cf_name: %s", name.c_str());
cf_list.emplace_back(name, ColumnFamilyOptions(options_));
}
s = versions.Recover(cf_list, true);
}
if (s.ok()) {
s = versions.GetLiveFilesChecksumInfo(checksum_list.get());
}
if (!s.ok()) {
return s;
}
// Step 3 verify the checksum
if (live_files.size() != checksum_list->size()) {
return Status::Corruption("The number of files does not match!");
}
for (size_t i = 0; i < live_files.size(); i++) {
std::string stored_checksum;
std::string stored_func_name;
s = checksum_list->SearchOneFileChecksum(
live_files[i].file_number, &stored_checksum, &stored_func_name);
if (s.IsNotFound()) {
return s;
}
if (live_files[i].file_checksum != stored_checksum ||
live_files[i].file_checksum_func_name != stored_func_name) {
return Status::Corruption(
"Checksum does not match! The file: " +
std::to_string(live_files[i].file_number) +
". In Manifest, checksum name: " + stored_func_name +
" and checksum " + stored_checksum +
". However, expected checksum name: " +
live_files[i].file_checksum_func_name + " and checksum " +
live_files[i].file_checksum);
}
}
return Status::OK();
}
// Verify the checksum of each file by recalculting the checksum and
// comparing it with the one being generated when a SST file is created.
Status VerifyEachFileChecksum() {
assert(db_ != nullptr);
EXPECT_OK(db_->DisableFileDeletions());
std::vector<LiveFileMetaData> live_files;
db_->GetLiveFilesMetaData(&live_files);
Status cs;
for (auto a_file : live_files) {
cs = VerifyChecksum(a_file);
if (!cs.ok()) {
break;
}
}
EXPECT_OK(db_->EnableFileDeletions(/*force=*/false));
return cs;
}
};
TEST_F(LdbCmdTest, DumpFileChecksumNoChecksum) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
FlushOptions fopts;
fopts.wait = true;
Random rnd(test::RandomSeed());
for (int i = 0; i < 200; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 100; i < 300; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 200; i < 400; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 300; i < 400; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->Close());
delete db;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "file_checksum_dump";
char arg4[] = "--hex";
char* argv[] = {arg1, arg2, arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify each sst file checksum value and checksum name
FileChecksumTestHelper fct_helper(opts, db, dbname);
ASSERT_OK(fct_helper.VerifyEachFileChecksum());
// Manually trigger compaction
char b_buf[16];
snprintf(b_buf, sizeof(b_buf), "%08d", 0);
char e_buf[16];
snprintf(e_buf, sizeof(e_buf), "%08d", 399);
Slice begin(b_buf);
Slice end(e_buf);
CompactRangeOptions options;
ASSERT_OK(db->CompactRange(options, &begin, &end));
// Verify each sst file checksum after compaction
FileChecksumTestHelper fct_helper_ac(opts, db, dbname);
ASSERT_OK(fct_helper_ac.VerifyEachFileChecksum());
ASSERT_OK(db->Close());
delete db;
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify the checksum information in memory is the same as that in Manifest;
std::vector<LiveFileMetaData> live_files;
db->GetLiveFilesMetaData(&live_files);
delete db;
ASSERT_OK(fct_helper_ac.VerifyChecksumInManifest(live_files));
}
TEST_F(LdbCmdTest, BlobDBDumpFileChecksumNoChecksum) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
opts.enable_blob_files = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
FlushOptions fopts;
fopts.wait = true;
Random rnd(test::RandomSeed());
for (int i = 0; i < 200; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 100; i < 300; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 200; i < 400; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 300; i < 400; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->Close());
delete db;
char arg1[] = "./ldb";
std::string arg2_str = "--db=" + dbname;
char arg3[] = "file_checksum_dump";
char arg4[] = "--hex";
char* argv[] = {arg1, const_cast<char*>(arg2_str.c_str()), arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify each sst and blob file checksum value and checksum name
FileChecksumTestHelper fct_helper(opts, db, dbname);
ASSERT_OK(fct_helper.VerifyEachFileChecksum());
// Manually trigger compaction
std::ostringstream oss_b_buf;
oss_b_buf << std::setfill('0') << std::setw(8) << std::fixed << 0;
std::ostringstream oss_e_buf;
oss_e_buf << std::setfill('0') << std::setw(8) << std::fixed << 399;
std::string b_buf = oss_b_buf.str();
std::string e_buf = oss_e_buf.str();
Slice begin(b_buf);
Slice end(e_buf);
CompactRangeOptions options;
ASSERT_OK(db->CompactRange(options, &begin, &end));
// Verify each sst file checksum after compaction
FileChecksumTestHelper fct_helper_ac(opts, db, dbname);
ASSERT_OK(fct_helper_ac.VerifyEachFileChecksum());
ASSERT_OK(db->Close());
delete db;
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
}
TEST_F(LdbCmdTest, DumpFileChecksumCRC32) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
opts.file_checksum_gen_factory = GetFileChecksumGenCrc32cFactory();
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
FlushOptions fopts;
fopts.wait = true;
Random rnd(test::RandomSeed());
for (int i = 0; i < 100; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 50; i < 150; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 100; i < 200; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 150; i < 250; i++) {
char buf[16];
snprintf(buf, sizeof(buf), "%08d", i);
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, buf, v));
}
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->Close());
delete db;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "file_checksum_dump";
char arg4[] = "--hex";
char* argv[] = {arg1, arg2, arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify each sst file checksum value and checksum name
FileChecksumTestHelper fct_helper(opts, db, dbname);
ASSERT_OK(fct_helper.VerifyEachFileChecksum());
// Manually trigger compaction
char b_buf[16];
snprintf(b_buf, sizeof(b_buf), "%08d", 0);
char e_buf[16];
snprintf(e_buf, sizeof(e_buf), "%08d", 249);
Slice begin(b_buf);
Slice end(e_buf);
CompactRangeOptions options;
ASSERT_OK(db->CompactRange(options, &begin, &end));
// Verify each sst file checksum after compaction
FileChecksumTestHelper fct_helper_ac(opts, db, dbname);
ASSERT_OK(fct_helper_ac.VerifyEachFileChecksum());
ASSERT_OK(db->Close());
delete db;
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify the checksum information in memory is the same as that in Manifest;
std::vector<LiveFileMetaData> live_files;
db->GetLiveFilesMetaData(&live_files);
ASSERT_OK(fct_helper_ac.VerifyChecksumInManifest(live_files));
ASSERT_OK(db->Close());
delete db;
}
TEST_F(LdbCmdTest, BlobDBDumpFileChecksumCRC32) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
opts.file_checksum_gen_factory = GetFileChecksumGenCrc32cFactory();
opts.enable_blob_files = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
FlushOptions fopts;
fopts.wait = true;
Random rnd(test::RandomSeed());
for (int i = 0; i < 100; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 50; i < 150; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 100; i < 200; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
for (int i = 150; i < 250; i++) {
std::ostringstream oss;
oss << std::setfill('0') << std::setw(8) << std::fixed << i;
std::string v = rnd.RandomString(100);
ASSERT_OK(db->Put(wopts, oss.str(), v));
}
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->Close());
delete db;
char arg1[] = "./ldb";
std::string arg2_str = "--db=" + dbname;
char arg3[] = "file_checksum_dump";
char arg4[] = "--hex";
char* argv[] = {arg1, const_cast<char*>(arg2_str.c_str()), arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, &db));
// Verify each sst and blob file checksum value and checksum name
FileChecksumTestHelper fct_helper(opts, db, dbname);
ASSERT_OK(fct_helper.VerifyEachFileChecksum());
// Manually trigger compaction
std::ostringstream oss_b_buf;
oss_b_buf << std::setfill('0') << std::setw(8) << std::fixed << 0;
std::ostringstream oss_e_buf;
oss_e_buf << std::setfill('0') << std::setw(8) << std::fixed << 249;
std::string b_buf = oss_b_buf.str();
std::string e_buf = oss_e_buf.str();
Slice begin(b_buf);
Slice end(e_buf);
CompactRangeOptions options;
ASSERT_OK(db->CompactRange(options, &begin, &end));
// Verify each sst file checksum after compaction
FileChecksumTestHelper fct_helper_ac(opts, db, dbname);
ASSERT_OK(fct_helper_ac.VerifyEachFileChecksum());
ASSERT_OK(db->Close());
delete db;
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
}
TEST_F(LdbCmdTest, OptionParsing) {
// test parsing flags
Options opts;
opts.env = TryLoadCustomOrDefaultEnv();
{
std::vector<std::string> args;
args.emplace_back("scan");
args.emplace_back("--ttl");
args.emplace_back("--timestamp");
LDBCommand* command = ROCKSDB_NAMESPACE::LDBCommand::InitFromCmdLineArgs(
args, opts, LDBOptions(), nullptr);
const std::vector<std::string> flags = command->TEST_GetFlags();
EXPECT_EQ(flags.size(), 2);
EXPECT_EQ(flags[0], "ttl");
EXPECT_EQ(flags[1], "timestamp");
delete command;
}
// test parsing options which contains equal sign in the option value
{
std::vector<std::string> args;
args.emplace_back("scan");
args.emplace_back("--db=/dev/shm/ldbtest/");
args.emplace_back(
"--from='abcd/efg/hijk/lmn/"
"opq:__rst.uvw.xyz?a=3+4+bcd+efghi&jk=lm_no&pq=rst-0&uv=wx-8&yz=a&bcd_"
"ef=gh.ijk'");
LDBCommand* command = ROCKSDB_NAMESPACE::LDBCommand::InitFromCmdLineArgs(
args, opts, LDBOptions(), nullptr);
const std::map<std::string, std::string> option_map =
command->TEST_GetOptionMap();
EXPECT_EQ(option_map.at("db"), "/dev/shm/ldbtest/");
EXPECT_EQ(option_map.at("from"),
"'abcd/efg/hijk/lmn/"
"opq:__rst.uvw.xyz?a=3+4+bcd+efghi&jk=lm_no&pq=rst-0&uv=wx-8&yz="
"a&bcd_ef=gh.ijk'");
delete command;
}
}
TEST_F(LdbCmdTest, ListFileTombstone) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
ASSERT_OK(db->Put(wopts, "foo", "1"));
ASSERT_OK(db->Put(wopts, "bar", "2"));
FlushOptions fopts;
fopts.wait = true;
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->DeleteRange(wopts, db->DefaultColumnFamily(), "foo", "foo2"));
ASSERT_OK(db->DeleteRange(wopts, db->DefaultColumnFamily(), "bar", "foo2"));
ASSERT_OK(db->Flush(fopts));
delete db;
{
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "list_file_range_deletes";
char* argv[] = {arg1, arg2, arg3};
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
"ListFileRangeDeletesCommand::DoCommand:BeforePrint", [&](void* arg) {
std::string* out_str = reinterpret_cast<std::string*>(arg);
// Count number of tombstones printed
int num_tb = 0;
const std::string kFingerprintStr = "start: ";
auto offset = out_str->find(kFingerprintStr);
while (offset != std::string::npos) {
num_tb++;
offset =
out_str->find(kFingerprintStr, offset + kFingerprintStr.size());
}
EXPECT_EQ(2, num_tb);
});
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(3, argv, opts, LDBOptions(), nullptr));
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
}
// Test the case of limiting tombstones
{
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "list_file_range_deletes";
char arg4[] = "--max_keys=1";
char* argv[] = {arg1, arg2, arg3, arg4};
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
"ListFileRangeDeletesCommand::DoCommand:BeforePrint", [&](void* arg) {
std::string* out_str = reinterpret_cast<std::string*>(arg);
// Count number of tombstones printed
int num_tb = 0;
const std::string kFingerprintStr = "start: ";
auto offset = out_str->find(kFingerprintStr);
while (offset != std::string::npos) {
num_tb++;
offset =
out_str->find(kFingerprintStr, offset + kFingerprintStr.size());
}
EXPECT_EQ(1, num_tb);
});
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
}
}
TEST_F(LdbCmdTest, DisableConsistencyChecks) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
{
DB* db = nullptr;
ASSERT_OK(DB::Open(opts, dbname, &db));
WriteOptions wopts;
FlushOptions fopts;
fopts.wait = true;
ASSERT_OK(db->Put(wopts, "foo1", "1"));
ASSERT_OK(db->Put(wopts, "bar1", "2"));
ASSERT_OK(db->Flush(fopts));
ASSERT_OK(db->Put(wopts, "foo2", "3"));
ASSERT_OK(db->Put(wopts, "bar2", "4"));
ASSERT_OK(db->Flush(fopts));
delete db;
}
{
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "checkconsistency";
char* argv[] = {arg1, arg2, arg3};
SyncPoint::GetInstance()->SetCallBack(
Clean up VersionStorageInfo a bit (#9494) Summary: The patch does some cleanup in and around `VersionStorageInfo`: * Renames the method `PrepareApply` to `PrepareAppend` in `Version` to make it clear that it is to be called before appending the `Version` to `VersionSet` (via `AppendVersion`), not before applying any `VersionEdit`s. * Introduces a helper method `VersionStorageInfo::PrepareForVersionAppend` (called by `Version::PrepareAppend`) that encapsulates the population of the various derived data structures in `VersionStorageInfo`, and turns the methods computing the derived structures (`UpdateNumNonEmptyLevels`, `CalculateBaseBytes` etc.) into private helpers. * Changes `Version::PrepareAppend` so it only calls `UpdateAccumulatedStats` if the `update_stats` flag is set. (Earlier, this was checked by the callee.) Related to this, it also moves the call to `ComputeCompensatedSizes` to `VersionStorageInfo::PrepareForVersionAppend`. * Updates and cleans up `version_builder_test`, `version_set_test`, and `compaction_picker_test` so `PrepareForVersionAppend` is called anytime a new `VersionStorageInfo` is set up or saved. This cleanup also involves splitting `VersionStorageInfoTest.MaxBytesForLevelDynamic` into multiple smaller test cases. * Fixes up a bunch of comments that were outdated or just plain incorrect. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9494 Test Plan: Ran `make check` and the crash test script for a while. Reviewed By: riversand963 Differential Revision: D33971666 Pulled By: ltamasi fbshipit-source-id: fda52faac7783041126e4f8dec0fe01bdcadf65a
2022-02-04 16:18:18 +00:00
"Version::PrepareAppend:forced_check", [&](void* arg) {
bool* forced = reinterpret_cast<bool*>(arg);
ASSERT_TRUE(*forced);
});
SyncPoint::GetInstance()->EnableProcessing();
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(3, argv, opts, LDBOptions(), nullptr));
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->DisableProcessing();
}
{
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "scan";
char* argv[] = {arg1, arg2, arg3};
SyncPoint::GetInstance()->SetCallBack(
Clean up VersionStorageInfo a bit (#9494) Summary: The patch does some cleanup in and around `VersionStorageInfo`: * Renames the method `PrepareApply` to `PrepareAppend` in `Version` to make it clear that it is to be called before appending the `Version` to `VersionSet` (via `AppendVersion`), not before applying any `VersionEdit`s. * Introduces a helper method `VersionStorageInfo::PrepareForVersionAppend` (called by `Version::PrepareAppend`) that encapsulates the population of the various derived data structures in `VersionStorageInfo`, and turns the methods computing the derived structures (`UpdateNumNonEmptyLevels`, `CalculateBaseBytes` etc.) into private helpers. * Changes `Version::PrepareAppend` so it only calls `UpdateAccumulatedStats` if the `update_stats` flag is set. (Earlier, this was checked by the callee.) Related to this, it also moves the call to `ComputeCompensatedSizes` to `VersionStorageInfo::PrepareForVersionAppend`. * Updates and cleans up `version_builder_test`, `version_set_test`, and `compaction_picker_test` so `PrepareForVersionAppend` is called anytime a new `VersionStorageInfo` is set up or saved. This cleanup also involves splitting `VersionStorageInfoTest.MaxBytesForLevelDynamic` into multiple smaller test cases. * Fixes up a bunch of comments that were outdated or just plain incorrect. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9494 Test Plan: Ran `make check` and the crash test script for a while. Reviewed By: riversand963 Differential Revision: D33971666 Pulled By: ltamasi fbshipit-source-id: fda52faac7783041126e4f8dec0fe01bdcadf65a
2022-02-04 16:18:18 +00:00
"Version::PrepareAppend:forced_check", [&](void* arg) {
bool* forced = reinterpret_cast<bool*>(arg);
ASSERT_TRUE(*forced);
});
SyncPoint::GetInstance()->EnableProcessing();
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(3, argv, opts, LDBOptions(), nullptr));
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->DisableProcessing();
}
{
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "scan";
char arg4[] = "--disable_consistency_checks";
char* argv[] = {arg1, arg2, arg3, arg4};
SyncPoint::GetInstance()->SetCallBack(
"ColumnFamilyData::ColumnFamilyData", [&](void* arg) {
ColumnFamilyOptions* cfo =
reinterpret_cast<ColumnFamilyOptions*>(arg);
ASSERT_FALSE(cfo->force_consistency_checks);
});
SyncPoint::GetInstance()->EnableProcessing();
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->DisableProcessing();
}
}
TEST_F(LdbCmdTest, TestBadDbPath) {
Env* base_env = TryLoadCustomOrDefaultEnv();
std::unique_ptr<Env> env(NewMemEnv(base_env));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s/.no_such_dir", dbname.c_str());
char arg3[1024];
snprintf(arg3, sizeof(arg3), "create_column_family");
char arg4[] = "bad cf";
char* argv[] = {arg1, arg2, arg3, arg4};
ASSERT_EQ(1,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
snprintf(arg3, sizeof(arg3), "drop_column_family");
ASSERT_EQ(1,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
}
namespace {
class WrappedEnv : public EnvWrapper {
public:
explicit WrappedEnv(Env* t) : EnvWrapper(t) {}
static const char* kClassName() { return "WrappedEnv"; }
const char* Name() const override { return kClassName(); }
};
} // namespace
TEST_F(LdbCmdTest, LoadCFOptionsAndOverride) {
// Env* base_env = TryLoadCustomOrDefaultEnv();
// std::unique_ptr<Env> env(NewMemEnv(base_env));
std::unique_ptr<Env> env(new WrappedEnv(Env::Default()));
Options opts;
opts.env = env.get();
opts.create_if_missing = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DestroyDB(dbname, opts));
ASSERT_OK(DB::Open(opts, dbname, &db));
ColumnFamilyHandle* cf_handle;
ColumnFamilyOptions cf_opts;
cf_opts.num_levels = 20;
ASSERT_OK(db->CreateColumnFamily(cf_opts, "cf1", &cf_handle));
delete cf_handle;
delete db;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "put";
char arg4[] = "key1";
char arg5[] = "value1";
char arg6[] = "--try_load_options";
char arg7[] = "--column_family=cf1";
char arg8[] = "--write_buffer_size=268435456";
char* argv[] = {arg1, arg2, arg3, arg4, arg5, arg6, arg7, arg8};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(8, argv, opts, LDBOptions(), nullptr));
ConfigOptions config_opts;
Options options;
std::vector<ColumnFamilyDescriptor> column_families;
config_opts.env = env.get();
ASSERT_OK(LoadLatestOptions(config_opts, dbname, &options, &column_families));
ASSERT_EQ(column_families.size(), 2);
ASSERT_EQ(options.num_levels, opts.num_levels);
ASSERT_EQ(column_families[1].options.num_levels, cf_opts.num_levels);
ASSERT_EQ(column_families[1].options.write_buffer_size, 268435456);
}
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
TEST_F(LdbCmdTest, UnsafeRemoveSstFile) {
Options opts;
opts.level0_file_num_compaction_trigger = 10;
opts.create_if_missing = true;
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(Env::Default(), "ldb_cmd_test");
ASSERT_OK(DestroyDB(dbname, opts));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
ASSERT_OK(DB::Open(opts, dbname, &db));
// Create three SST files
for (size_t i = 0; i < 3; ++i) {
ASSERT_OK(db->Put(WriteOptions(), std::to_string(i), std::to_string(i)));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
ASSERT_OK(db->Flush(FlushOptions()));
}
// Determine which is the "middle" one
std::vector<LiveFileMetaData> sst_files;
db->GetLiveFilesMetaData(&sst_files);
std::vector<uint64_t> numbers;
for (auto& f : sst_files) {
numbers.push_back(f.file_number);
}
ASSERT_EQ(numbers.size(), 3);
std::sort(numbers.begin(), numbers.end());
uint64_t to_remove = numbers[1];
// Close for unsafe_remove_sst_file
delete db;
db = nullptr;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "unsafe_remove_sst_file";
char arg4[20];
snprintf(arg4, sizeof(arg4), "%" PRIu64, to_remove);
char* argv[] = {arg1, arg2, arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
// Re-open, and verify with Get that middle file is gone
ASSERT_OK(DB::Open(opts, dbname, &db));
std::string val;
ASSERT_OK(db->Get(ReadOptions(), "0", &val));
ASSERT_EQ(val, "0");
ASSERT_OK(db->Get(ReadOptions(), "2", &val));
ASSERT_EQ(val, "2");
ASSERT_TRUE(db->Get(ReadOptions(), "1", &val).IsNotFound());
// Now with extra CF, two more files
ColumnFamilyHandle* cf_handle;
ColumnFamilyOptions cf_opts;
ASSERT_OK(db->CreateColumnFamily(cf_opts, "cf1", &cf_handle));
for (size_t i = 3; i < 5; ++i) {
ASSERT_OK(db->Put(WriteOptions(), cf_handle, std::to_string(i),
std::to_string(i)));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
ASSERT_OK(db->Flush(FlushOptions(), cf_handle));
}
// Determine which is the "last" one
sst_files.clear();
db->GetLiveFilesMetaData(&sst_files);
numbers.clear();
for (auto& f : sst_files) {
numbers.push_back(f.file_number);
}
ASSERT_EQ(numbers.size(), 4);
std::sort(numbers.begin(), numbers.end());
to_remove = numbers.back();
// Close for unsafe_remove_sst_file
delete cf_handle;
delete db;
db = nullptr;
snprintf(arg4, sizeof(arg4), "%" PRIu64, to_remove);
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
std::vector<ColumnFamilyDescriptor> cfds = {{kDefaultColumnFamilyName, opts},
{"cf1", cf_opts}};
std::vector<ColumnFamilyHandle*> handles;
ASSERT_OK(DB::Open(opts, dbname, cfds, &handles, &db));
ASSERT_OK(db->Get(ReadOptions(), handles[1], "3", &val));
ASSERT_EQ(val, "3");
ASSERT_TRUE(db->Get(ReadOptions(), handles[1], "4", &val).IsNotFound());
ASSERT_OK(db->Get(ReadOptions(), handles[0], "0", &val));
ASSERT_EQ(val, "0");
// Determine which is the "first" one (most likely to be opened in recovery)
sst_files.clear();
db->GetLiveFilesMetaData(&sst_files);
numbers.clear();
for (auto& f : sst_files) {
numbers.push_back(f.file_number);
}
ASSERT_EQ(numbers.size(), 3);
std::sort(numbers.begin(), numbers.end());
to_remove = numbers.front();
// This time physically delete the file before unsafe_remove
{
std::string f = dbname + "/" + MakeTableFileName(to_remove);
ASSERT_OK(Env::Default()->DeleteFile(f));
}
// Close for unsafe_remove_sst_file
for (auto& h : handles) {
delete h;
}
delete db;
db = nullptr;
snprintf(arg4, sizeof(arg4), "%" PRIu64, to_remove);
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
ASSERT_OK(DB::Open(opts, dbname, cfds, &handles, &db));
ASSERT_OK(db->Get(ReadOptions(), handles[1], "3", &val));
ASSERT_EQ(val, "3");
ASSERT_TRUE(db->Get(ReadOptions(), handles[0], "0", &val).IsNotFound());
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
for (auto& h : handles) {
delete h;
}
delete db;
}
TEST_F(LdbCmdTest, FileTemperatureUpdateManifest) {
auto test_fs = std::make_shared<FileTemperatureTestFS>(FileSystem::Default());
std::unique_ptr<Env> env(new CompositeEnvWrapper(Env::Default(), test_fs));
Options opts;
opts.bottommost_temperature = Temperature::kWarm;
opts.level0_file_num_compaction_trigger = 10;
opts.create_if_missing = true;
opts.env = env.get();
DB* db = nullptr;
std::string dbname = test::PerThreadDBPath(env.get(), "ldb_cmd_test");
ASSERT_OK(DestroyDB(dbname, opts));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
ASSERT_OK(DB::Open(opts, dbname, &db));
std::array<Temperature, 5> kTestTemps = {
Temperature::kCold, Temperature::kWarm, Temperature::kHot,
Temperature::kWarm, Temperature::kCold};
std::map<uint64_t, Temperature> number_to_temp;
for (size_t i = 0; i < kTestTemps.size(); ++i) {
ASSERT_OK(db->Put(WriteOptions(), std::to_string(i), std::to_string(i)));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
ASSERT_OK(db->Flush(FlushOptions()));
std::map<uint64_t, Temperature> current_temps;
test_fs->CopyCurrentSstFileTemperatures(&current_temps);
for (auto e : current_temps) {
if (e.second == Temperature::kUnknown) {
test_fs->OverrideSstFileTemperature(e.first, kTestTemps[i]);
number_to_temp[e.first] = kTestTemps[i];
}
}
}
// Close & reopen
delete db;
db = nullptr;
test_fs->PopRequestedSstFileTemperatures();
ASSERT_OK(DB::Open(opts, dbname, &db));
for (size_t i = 0; i < kTestTemps.size(); ++i) {
std::string val;
ASSERT_OK(db->Get(ReadOptions(), std::to_string(i), &val));
ASSERT_EQ(val, std::to_string(i));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
}
// Still all unknown
std::vector<std::pair<uint64_t, Temperature>> requests;
test_fs->PopRequestedSstFileTemperatures(&requests);
ASSERT_EQ(requests.size(), kTestTemps.size());
for (auto& r : requests) {
ASSERT_EQ(r.second, Temperature::kUnknown);
}
// Close for update_manifest
delete db;
db = nullptr;
char arg1[] = "./ldb";
char arg2[1024];
snprintf(arg2, sizeof(arg2), "--db=%s", dbname.c_str());
char arg3[] = "update_manifest";
char arg4[] = "--update_temperatures";
char* argv[] = {arg1, arg2, arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), nullptr));
// Re-open, get, and verify manifest temps (based on request)
test_fs->PopRequestedSstFileTemperatures();
ASSERT_OK(DB::Open(opts, dbname, &db));
for (size_t i = 0; i < kTestTemps.size(); ++i) {
std::string val;
ASSERT_OK(db->Get(ReadOptions(), std::to_string(i), &val));
ASSERT_EQ(val, std::to_string(i));
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
}
requests.clear();
test_fs->PopRequestedSstFileTemperatures(&requests);
ASSERT_EQ(requests.size(), kTestTemps.size());
for (auto& r : requests) {
ASSERT_EQ(r.second, number_to_temp[r.first]);
}
delete db;
}
TEST_F(LdbCmdTest, RenameDbAndLoadOptions) {
Env* env = TryLoadCustomOrDefaultEnv();
Options opts;
opts.env = env;
opts.create_if_missing = false;
std::string old_dbname = test::PerThreadDBPath(env, "ldb_cmd_test");
std::string new_dbname = old_dbname + "_2";
ASSERT_OK(DestroyDB(old_dbname, opts));
ASSERT_OK(DestroyDB(new_dbname, opts));
char old_arg[1024];
snprintf(old_arg, sizeof(old_arg), "--db=%s", old_dbname.c_str());
char new_arg[1024];
snprintf(new_arg, sizeof(old_arg), "--db=%s", new_dbname.c_str());
const char* argv1[] = {"./ldb",
old_arg,
"put",
"key1",
"value1",
"--try_load_options",
"--create_if_missing"};
const char* argv2[] = {"./ldb", old_arg, "get", "key1", "--try_load_options"};
const char* argv3[] = {"./ldb", new_arg, "put",
"key2", "value2", "--try_load_options"};
const char* argv4[] = {"./ldb", new_arg, "get", "key1", "--try_load_options"};
const char* argv5[] = {"./ldb", new_arg, "get", "key2", "--try_load_options"};
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(7, argv1, opts, LDBOptions(), nullptr));
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(5, argv2, opts, LDBOptions(), nullptr));
ConfigOptions config_opts;
Options options;
std::vector<ColumnFamilyDescriptor> column_families;
config_opts.env = env;
ASSERT_OK(
LoadLatestOptions(config_opts, old_dbname, &options, &column_families));
ASSERT_EQ(options.wal_dir, "");
ASSERT_OK(env->RenameFile(old_dbname, new_dbname));
ASSERT_NE(
0, LDBCommandRunner::RunCommand(6, argv1, opts, LDBOptions(), nullptr));
ASSERT_NE(
0, LDBCommandRunner::RunCommand(5, argv2, opts, LDBOptions(), nullptr));
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(6, argv3, opts, LDBOptions(), nullptr));
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(5, argv4, opts, LDBOptions(), nullptr));
ASSERT_EQ(
0, LDBCommandRunner::RunCommand(5, argv5, opts, LDBOptions(), nullptr));
ASSERT_OK(DestroyDB(new_dbname, opts));
}
Add manifest fix-up utility for file temperatures (#9683) Summary: The goal of this change is to allow changes to the "current" (in FileSystem) file temperatures to feed back into DB metadata, so that they can inform decisions and stats reporting. In part because of modular code factoring, it doesn't seem easy to do this automagically, where opening an SST file and observing current Temperature different from expected would trigger a change in metadata and DB manifest write (essentially giving the deep read path access to the write path). It is also difficult to do this while the DB is open because of the limitations of LogAndApply. This change allows updating file temperature metadata on a closed DB using an experimental utility function UpdateManifestForFilesState() or `ldb update_manifest --update_temperatures`. This should suffice for "migration" scenarios where outside tooling has placed or re-arranged DB files into a (different) tiered configuration without going through RocksDB itself (currently, only compaction can change temperature metadata). Some details: * Refactored and added unit test for `ldb unsafe_remove_sst_file` because of shared functionality * Pulled in autovector.h changes from https://github.com/facebook/rocksdb/issues/9546 to fix SuperVersionContext move constructor (related to an older draft of this change) Possible follow-up work: * Support updating manifest with file checksums, such as when a new checksum function is used and want existing DB metadata updated for it. * It's possible that for some repair scenarios, lighter weight than full repair, we might want to support UpdateManifestForFilesState() to modify critical file details like size or checksum using same algorithm. But let's make sure these are differentiated from modifying file details in ways that don't suspect corruption (or require extreme trust). Pull Request resolved: https://github.com/facebook/rocksdb/pull/9683 Test Plan: unit tests added Reviewed By: jay-zhuang Differential Revision: D34798828 Pulled By: pdillinger fbshipit-source-id: cfd83e8fb10761d8c9e7f9c020d68c9106a95554
2022-03-18 23:35:51 +00:00
class MyComparator : public Comparator {
public:
int Compare(const Slice& a, const Slice& b) const override {
return a.compare(b);
}
void FindShortSuccessor(std::string* /*key*/) const override {}
void FindShortestSeparator(std::string* /*start*/,
const Slice& /*limit*/) const override {}
const char* Name() const override { return "my_comparator"; }
};
TEST_F(LdbCmdTest, CustomComparator) {
Env* env = TryLoadCustomOrDefaultEnv();
MyComparator my_comparator;
Options opts;
opts.env = env;
opts.create_if_missing = true;
opts.create_missing_column_families = true;
opts.comparator = &my_comparator;
std::string dbname = test::PerThreadDBPath(env, "ldb_cmd_test");
DB* db = nullptr;
std::vector<ColumnFamilyDescriptor> cfds = {
{kDefaultColumnFamilyName, opts}, {"cf1", opts}, {"cf2", opts}};
std::vector<ColumnFamilyHandle*> handles;
ASSERT_OK(DestroyDB(dbname, opts));
ASSERT_OK(DB::Open(opts, dbname, cfds, &handles, &db));
ASSERT_OK(db->Put(WriteOptions(), "k1", "v1"));
for (auto& h : handles) {
ASSERT_OK(db->DestroyColumnFamilyHandle(h));
}
delete db;
char arg1[] = "./ldb";
std::string arg2 = "--db=" + dbname;
char arg3[] = "get";
char arg4[] = "k1";
char* argv[] = {arg1, const_cast<char*>(arg2.c_str()), arg3, arg4};
ASSERT_EQ(0,
LDBCommandRunner::RunCommand(4, argv, opts, LDBOptions(), &cfds));
}
} // namespace ROCKSDB_NAMESPACE
int main(int argc, char** argv) {
ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
RegisterCustomObjects(argc, argv);
return RUN_ALL_TESTS();
}